Polynucleotide sequencing continues to find increasing use in medical applications such as genetic screening and genotyping of tumors. Many polynucleotide sequencing methods rely on sample processing techniques of the original sample, including random fragmentation of polynucleotides. These processing techniques can provide advantages in terms of throughput and efficiency, but the resultant sequence information obtained from these processed samples can lack important contextual information in terms of the location of particular sequences within the broader linear (two-dimensional) sequence of the original nucleic acid molecule that contained those sequences. Structural context within the three dimensional space of the original sample is also lost with many sample processing and sequencing techniques. There is thus a need for sequencing technologies that retain structural and molecular context of the identified nucleic acid sequences.